Premium
Knowing when to draw the line: designing more informative ecological experiments
Author(s) -
Cottingham Kathryn L.,
Len Jay T.,
Brown Bryan L.
Publication year - 2005
Publication title -
frontiers in ecology and the environment
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.918
H-Index - 164
eISSN - 1540-9309
pISSN - 1540-9295
DOI - 10.1890/1540-9295(2005)003[0145:kwtdtl]2.0.co;2
Subject(s) - analysis of variance , regression analysis , statistics , regression , variance (accounting) , linear regression , variables , variable (mathematics) , econometrics , mathematics , mathematical analysis , accounting , business
Linear regression and analysis of variance (ANOVA) are two of the most widely used statistical techniques in ecology. Regression quantitatively describes the relationship between a response variable and one or more continuous independent variables, while ANOVA determines whether a response variable differs among discrete values of the independent variable(s). Designing experiments with discrete factors is straightforward because ANOVA is the only option, but what is the best way to design experiments involving continuous factors? Should ecologists prefer experiments with few treatments and many replicates analyzed with ANOVA, or experiments with many treatments and few replicates per treatment analyzed with regression? We recommend that ecologists choose regression, especially replicated regression, over ANOVA when dealing with continuous factors for two reasons: (1) regression is generally a more powerful approach than ANOVA and (2) regression provides quantitative output that can be incorporated into ecological models more effectively than ANOVA output.